| title | PointDiT: Pixel-Space Diffusion for Monocular Geometry Estimation |
|---|---|
| authors | Haofei Xu, Rundi Wu, Philipp Henzler, Nikolai Kalischek, Michael Oechsle, Fabian Manhardt, Marc Pollefeys, Andreas Geiger, Federico Tombari, Michael Niemeyer |
| date | 2026-07-02 |
| venue | arXiv (preprint, not peer reviewed) |
| id | arXiv:2607.02515 |
| affiliation | |
| upvotes | 19 (Hugging Face Daily Papers, 2026-11-01) |
| links | abs pdf hf |
| abstract | State-of-the-art single-image 3D reconstruction methods often rely on complex hybrid architectures and loss functions, or compress geometry into latent spaces in order to leverage pre-trained latent diffusion models. In this work, we show that such architectural overhead and intricate loss formulations are unnecessary. We introduce a minimalist pixel-space Diffusion Transformer, built on a plain ViT, that operates directly on raw 3D point map patches and is conditioned on image tokens from a pre-trained DINOv3. Unlike existing latent diffusion approaches, we train our diffusion backbone entirely from scratch, eliminating the need for point map tokenizers. Despite its simplicity, our approach surpasses complex latent-based diffusion models while remaining significantly simpler than hybrid alternatives. Notably, it produces sharper geometric structure and is more robust in highly ambiguous regions, such as transparent objects. |
cite
@misc{xu2026pointdit,
title={PointDiT: Pixel-Space Diffusion for Monocular Geometry Estimation},
author={Haofei Xu and Rundi Wu and Philipp Henzler and Nikolai Kalischek and Michael Oechsle and Fabian Manhardt and Marc Pollefeys and Andreas Geiger and Federico Tombari and Michael Niemeyer},
year={2026},
eprint={2607.02515},
archivePrefix={arXiv},
url={https://arxiv.org/abs/2607.02515}
}